From 670551a4279fdc45c43b276d01633eeff43ccc9d Mon Sep 17 00:00:00 2001
From: Pompolic <pompolic@special-circumstanc.es>
Date: Thu, 13 Feb 2020 20:24:24 +0100
Subject: [PATCH] WIP: grammar definition for ASCIIHexDecode
---
pdf.c | 64 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++--
1 file changed, 62 insertions(+), 2 deletions(-)
diff --git a/pdf.c b/pdf.c
index 23cf121..3e4cb37 100644
--- a/pdf.c
+++ b/pdf.c
@@ -222,6 +222,21 @@ act_hupper(const HParseResult *p, void *u)
return H_MAKE_UINT(H_CAST_UINT(p->ast) - 'A');
}
+HParsedToken *
+act_ahextruncated(const HParseResult *p, void *u)
+{
+ uint8_t b = 0;
+ HCountedArray *seq = H_CAST_SEQ(p->ast);
+
+ /* Assumption: At this point seq->elements[0] is a hex digit
+ * and seq->elements[1] holds '>' (EOD)
+ */
+ // XXX figure out how to compare to '>'
+ assert(seq->used == 2);
+ b = H_CAST_UINT(seq->elements[0]) << 4;
+ return H_MAKE_UINT(b);
+}
+
HParsedToken *
act_nat(const HParseResult *p, void *u)
{
@@ -1043,6 +1058,47 @@ FlateDecode(const Dict *parms, HBytes b, HParser *p)
return res;
}
+/*
+ * Decodes ASCII hexadecimal data into binary data.
+ * parms should be empty, because the filter has no parameters
+ */
+HParseResult *
+ASCIIHexDecode(const Dict *parms, HBytes b, HParser *p)
+{
+ HParseResult *res;
+ H_RULE(lwchar, IN(LWCHARS));
+ H_ARULE(digit, h_ch_range('0', '9'));
+ H_ARULE(hlower, h_ch_range('a', 'f'));
+ H_ARULE(hupper, h_ch_range('A', 'F'));
+ H_RULE(ahexeod, h_ch('>'));
+ H_RULE(hdigit, CHX(digit, hlower, hupper));
+ H_RULE(hdigitpair, h_middle(SEQ(lwchar),SEQ(hdigit,hdigit),SEQ(lwchar)));
+ H_ARULE(ahextruncated, SEQ(hdigit,ahexeod));
+
+ H_RULE(hs_end, CHX(hdigitpair, ahextruncated));
+ H_RULE(hexstream, SEQ(h_many(hdigitpair),hs_end));
+
+ res = h_parse(hexstream, b.token, b.len);
+ if(!res)
+ {
+ fprintf(stderr, "parse error in ASCIIHexDecode filter\n");
+ return NULL;
+ }
+
+ return res;
+}
+
+/*
+ * Decodes ASCII base-85 encoded data and produces binary data.
+ * parms should be empty, because the filter has no parameters
+ */
+HParseResult*
+ASCII85Decode(const Dict *parms, HBytes b, HParser *p)
+{
+ fprintf(stderr, "ASCII85Decode: not implemented\n");
+ return NULL;
+}
+
/*
* decode the bytes in 'b' according to metadata in the stream dictionary 'd'
* and parse the result with 'p'.
@@ -1066,11 +1122,15 @@ decode_stream(const Dict *d, HBytes b, HParser *p)
return NULL; // XXX filter chains not supported, yet
assert(v->token_type == TT_BYTES);
if (bytes_eq(v->bytes, "FlateDecode"))
- filter = FlateDecode;
+ filter = FlateDecode; // XXX add ASCIIHexDecode and ASCII85Decode here
+ else if (bytes_eq(v->bytes, "ASCIIHexDecode"))
+ filter = ASCIIHexDecode;
+ else if (bytes_eq(v->bytes, "ASCII85Decode"))
+ filter = ASCII85Decode;
else
return NULL; /* filter not supported */
- v = dictentry(d, "DecodeParms");
+ v = dictentry(d, "DecodeParms"); // XXX ASCII filters don't use DecodeParms
if (v && v->token_type == TT_Dict)
parms = v->user;
--
GitLab